# lec1 - Data Analysis for Financial Engineers Professor S...

This preview shows pages 1–5. Sign up to view the full content.

Data Analysis for Financial Engineers Professor S. Kou Department of IEOR, Columbia University Lecture 1a. Introduction 1 What is this course about? Data Analysis for FE: Time Series (AR, MA, ARMA, ARCH, GARCH) Regression (Time series regression) Empirical Finance. (Capital Asset Pricing Model, fac- tor Models, e cient market hypothesis, continuous time models) 2 Applications. 1. Hedging & Risk Management 2. Program Trading In the classical Black-Scholes model, it is assumed that the dynamics of the stock prices is given by ( )= (0) exp {  +  ( ) } where ( ) is a standard Brownian motion. 3 After we have such wonderful models, why do we need to learn Data Analysis? Three reasons: (1) Test whether models f t data. (2) Estimate parameters in the models. (3) Build new Mathematical/Statistical models. For exam- ple, alternative models to the Black-Scholes model. 4

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
Dow Jones Index On Oct 19, 1987, Monday, the DJ Industrial Average fell by 25.6%. From Jan 1, 1980 to Oct 16, 1987, Friday, the standard deviation of daily returns on the Dow was about 1.16%. The drop of 25.6% was about a negative return of 22 standard deviations, after re-centering the mean of the daily return (which is close to zero). 5 If the Black-scholes model is correct, then the stock return will have a normal distribution. ( ≤− 22) = Φ ( 22) = 1 4 10 107 The universe is believed to have existed for 15 billion years or about 5 10 17 seconds. 6 Mathematics will be used in this course includes: 1. Intro to Statistics. 2. Some optimization tools. 3. S-plus programming language 7 What are S-Plus and S+FinMetrics? S was originally developed at Bell Labs. S was bought by Insightful Corporation and became S-Plus. FinMetrics was added to S-Plus around 2000. It mainly deals with f nancial time series, including GARCH models and time series regression. 8
Students are responsible to get the necessary software pack- ages. (Just like the textbook requirement). Inquire about how to get S-Plus and FinMetrics should be addressed to www.insightful.com Di f erences between S-Plus and other software packages, such as R, SAS, Matlab, etc. 9 Review of Basic Statistics 1. Random Variables Discrete: binomial, Poisson,. .. Continuous: normal, exponential, gamma, 2 ,. . . 10 S-plus commands: rnorm: generating N(0,1) rpois: generating Poisson random variables help(rnorm): ask for help in S-plus. normal density: S-plus commend: dnorm Cumulative distribution function (cdf): ( )= ( ) The cdf of N(0,1) is given by Φ ( Z −∞ ( )  11 S-plus: pnorm 12

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
2 , and distributions Suppose , =1 2   are standard normal random vari- ables. Then X =1 2 2 the chi-squared distribution with d.f. . = p 2  13 1  2 = 2 1  1 2 2  2 14 Expectation: discrete: ( )= P  ( = ) continuous: ( R −∞  ( )  Variance  ( [( ( )) 2 ]= [ 2 ] ( [ ]) 2 Standard deviation:  (
This is the end of the preview. Sign up to access the rest of the document.

## This note was uploaded on 10/16/2010 for the course IEOR 4709 taught by Professor Stevenkou during the Fall '10 term at Columbia.

### Page1 / 22

lec1 - Data Analysis for Financial Engineers Professor S...

This preview shows document pages 1 - 5. Sign up to view the full document.

View Full Document
Ask a homework question - tutors are online